Video-Eval Stage2: Annotation Refining

{{ before_start_anno | safe }}

Videos Gallery -- See examples in each sub-score

{{ visual_def | safe}}

Visual Quality - Good

Visual Quality - Avg

Visual Quality - Bad

{{ temporal_def | safe}}

Temporal Consistency - Good

Temporal Consistency - Avg

Temporal Consistency - Bad

{{ dynamic_def | safe}}

Dynamic Degree - Good

Dynamic Degree - Avg

Dynamic Degree - Bad

{{ align_def | safe}}

Good

Average: prompt is "A golden retriever puppy holding a green sign that says 'NVIDIA ROCKS'", but the slogan is not spelled correctly in the video, error point d, affecting video quality to some extent

Average: prompt is "A skull surrounded with smoke and fire", but fire is not expressed in the video, error point a, affecting video quality to some extent

Bad: prompt is "tiger coming out from TV", but no tiger, error point a, affecting video quality greatly

Bad: prompt is "a beautiful broadcaster", but it's not expressed in video at all, error point a, affecting video quality greatly

{{ factual_def | safe }}

Factual Consistency - Good

Factual Consistency - Avg

Factual Consistency - Bad